1 The UrbanShift project

1.1 Objectives

UrbanShift is a global program that supports cities to adopt integrated approaches to urban development, shaping low-carbon, climate-resilient communities where people and planet both can thrive. The global program is funded by the Global Environment Facility (GEF) led by the UN Environment Programme (UNEP), in partnership with the World Resources Institute (WRI), C40 Cities, ICLEI – Local Governments for Sustainability. The initiative supports 23 cities across nine countries, providing the knowledge, tools and training they need to transform their urban fabric and shift towards a more sustainable, equitable future.

As one of the key activities to support UrbanShift cities , the WRI data team will work with UrbanShift cities to identify and provide all cities with a common set of critical spatial data layers. using open source, global data. World Resources Institute is providing several types of data-related assistance to participating cities:

  • A suite of key geo-spatial layers
  • Baseline measurements of core UrbanShift indicators
  • Geo-spatial analysis on selected thematic areas
  • Capacity building and technical assistance on data governance and geospatial data as part of the City Academy and Labs modules of UrbanShift.

Outputs will include datasets, indicators and replicable analysis methods relevant to all cities. Additionally, analyses customized to the specific themes of interest for each city will be provided. Finally, an UrbanShift Lab will be delivered for which these data and analyses may act as one input.

1.2 Baseline indicators

To help understand the current status and identify changes of sustainability in UrbanShift cities, we aim to measure key baseline indicators for all cities using comparable approaches. The selected indicators focus on measuring the status and change of the core objectives of the global project, which are aligned with three of Global Environment Facility’s focal areas for its current investment cycle (GEF-7):land degradation, biodiversity, and greenhouse gas emissions.

These assessments are intended to support the evaluation of patterns within and between cities and to provide contextual information to cities to help them in the deision making process. We will disseminate the results to help local governments, the global project team, implementing agencies and national governments to gain a better understanding of the cities’ current status as it relates to sustainability efforts, capacities, needs and opportunities, and planned investments.

2 Land degradation & green space monitoring

2.1 Context

Land degradation is one of the world’s most pressing environmental challenge with direct impacts on climate change adaptation, ecosystem condition, food security and human well-being. Globally, about 25% of the total land area has been degraded and 3.2 billion people are affected by this phenomenon, particularly rural communities, smallholder farmers, and the very poor (source). As a financial mechanism of the United Nations Convention to Combat Desertification (UNCDD), the GEF is highlighting the necessity to invest in programs that encourage sustainable land management practices and land degradation has been selected as one of the strategic focal area in its new four-year investment cycle (known as GEF-7).

Through the Land Degradation Neutrality (LDN) program, the UNCDD, in collaboration with multiple international partners, is supporting interested countries in setting national baselines, targets and measures to protect their land resources. The Land Degradation Neutrality concept is defined as a state whereby the amount and quality of land resources necessary to support ecosystem function and services and enhance food security remain stable or increase within specified temporal and spatial scales and ecosystems.

These objectives are in accordance with the Sustainable Development Goal (SDG) target 15.3 stating: By 2030, combat desertification, restore degraded land and soil, including land affected by desertification, drought and floods, and strive to achieve a land degradation neutral world.

2.2 Definitions

The Special Report on Climate Change and Land, defines Land degradation as negative trend in land condition, caused by direct or indirect human-induced processes including anthropogenic climate change, expressed as long-term reduction or loss of at least one of the following: biological productivity, ecological integrity, or value to humans.

Multiple factors are increasing the pressure on land resources such as the growing demand for food, urban expansion, decrease in land productivity due to soil degradation, biodiversity loss and extreme weather events.

2.3 Data

In order to provide greenspace indicators within UrbanShift cities, we selected global coverage datasets with high spatial resolution.

  • Dynamic world (DW): The Dynamic World Land Cover product displays a global map of land use/land cover (LULC) provided from ESA Sentinel-2 imagery at 10m resolution. It is composed of 10 land use classes: water, trees,grass,flooded vegetation,crops,scrub/shrub, built area,bare ground, and snow/ice. The DW datasets can be used as a proxy for estimating land degradation by quantifying the percent of vegetation land area (such as water, trees,grass) within UrbanShift cities boundaries.

  • Sentinel-2: Sentinel-2 is a wide-swath, high-resolution, multi-spectral imaging mission supporting Copernicus Land Monitoring studies, including the monitoring of vegetation, soil and water cover. Its optical instrument samples in 13 spectral bands: four bands at 10 meters, six bands at 20 meters and three bands at 60 meters spatial resolution. Sentinel-2 images can be used for computing the Normalized Difference Vegetation Index (NDVI) considered as an effective index for estimating green vegetation.

  • Tree Outside of Forests (TOF): The TOF project provides tree extent data at 10m scale based on trained Convolutional Neural Network using satellite imagery (Sentinel-1 and Sentinel-2). It enables accurate reporting of tree cover outside of dense, closed-canopy forests and urban areas. For more details about the data, see the github repository and this article. The TOF data is used for estimating tree cover within the selected UrbanShift cities.

  • Intra-Urban Land Use (ULU): The ULU data provides land use and land cover information of urban areas based on the application of supervised classification model trained on high resolution Sentinel-2 satellite imagery data. Urban land classes include: open space,non residential area,residential atomistic,residential informal,residential forma,housing project, and `roads``. (detailed documentation of this dataset. This dataset provides the distribution of different urban land use classes within UrbanShift cities’ boundaries and statistics on vegetation and tree cover levels by land use classes

2.4 Greenspace indicators

Based on the previously identified datasets, we propose to compute a list of indicators that enable us to assess land degradation and greenspace status within UrbanShift cities. The table below lists the different indicators we are measuring in this analysis:

Indicator name Description Used datasets Years
Percent of vegetation land based on Dynamic World land cover classes Percent of land that is trees/water/grass/Scrub/flooded vegetation land cover Dynamic World Land cover [2016,2020]
Percent of vegetation land based on NDVI threshold Percent of land that is vegetation (NDVI threshold > 0.4) Sentinel-2 [2016,2020]
Percent of land with tree cover Percent of land that has tree cover Tree Outside of Forests (TOF) [2020]
Percent of built area with tree cover Percent of land that has tree cover Dynamic World Land cover, Tree Outside of Forests (TOF) [2020]
Percent of built area with vegetation Percent of built area with vegetation based on NDVI threshold Dynamic World Land cover, Sentinel-2 [2016,2020]
Percent of Intra-Urban land classes Percent of land based on Urban Land Use classification: Open space,Residential,Atomistic,Informal subdivision, Formal subdivision,Housing projects. Intra-Urban Land Use [2020]
Percent of tree cover by urban land classes Percent of tree cover level (as expressed in Tree Outside of Forests dataset) by intra-urbal land use classes Intra-Urban Land Use, Tree Outside of Forests (TOF) [2020]
Percent of vegetation by urban land classes Percent of vegetation (based on NDVI threshold) by intra-urbal land use classes Intra-Urban Land Use, Sentinel-2 [2020]

3 Case study

3.1 Administrative boundaries

The administrative boundaries data is obtained from the geoBoundaries database. Produced and maintained by the William & Mary geoLab and open data community since 2017, the geoBoundaries Global Database of Political Administrative Boundaries Database is an online, open license resource of boundaries (i.e., state, county) for every country in the world.

The administrative boundaries are used for extracting and aggregating geospatial information and indicators based on the city extent.

3.2 Percent of vegetation land (Dynamic World)

This Dynamic World Land Cover product displays a global map of land use/land cover (LULC) provided from ESA Sentinel-2 imagery at 10m resolution. It is composed of 10 land use classes based on a deep learning model. The class definition is as follows (Reference):

  • Water: Areas where water was predominantly present throughout the year; may not cover areas with sporadic or ephemeral water; contains little to no sparse vegetation, no rock outcrop nor built up features like docks; examples: rivers, ponds, lakes, oceans, flooded salt plains.
  • Trees: Any significant clustering of tall (~15-m or higher) dense vegetation, typically with a closed or dense canopy.
  • Grass: Open areas covered in homogenous grasses with little to no taller vegetation.
  • Flooded vegetation: Areas of any type of vegetation with obvious intermixing of water throughout a majority of the year.
  • Crops: Human planted/plotted cereals, grasses, and crops not at tree height; examples: corn, wheat, soy, fallow plots of structured land.
  • Scrub/shrub: Mix of small clusters of plants or single plants dispersed on a landscape that shows exposed soil or rock (examples: moderate to sparse cover of bushes, shrubs and tufts of grass, savannas with very sparse grasses, trees or other plants)
  • Built Area: Human made structures; major road and rail networks; large homogenous impervious surfaces including parking structures, office buildings and residential housing; examples: houses, dense villages / towns / cities, paved roads, asphalt.
  • Bare ground: Areas of rock or soil with very sparse to no vegetation for the entire year; large areas of sand and deserts with no to little vegetation; examples: exposed rock or soil, desert and sand dunes, dry salt flats/pans, dried lake beds, mines.
  • Snow/Ice: Large homogeneous areas of permanent snow or ice, typically only in mountain areas or highest latitudes; examples: glaciers, permanent snowpack, snow fields.

The map below displays the spatial distribution of Land cover for the selected city on 2020:

city_id indicator_theme data_sources indicator_name year value
RWA-Kigali greenspace Dynamic WOrld dynamic_world_vegetation_land_percent 2016 49.22
RWA-Kigali greenspace Dynamic WOrld dynamic_world_vegetation_land_percent 2017 44.15
RWA-Kigali greenspace Dynamic WOrld dynamic_world_vegetation_land_percent 2018 34.90
RWA-Kigali greenspace Dynamic WOrld dynamic_world_vegetation_land_percent 2019 39.11
RWA-Kigali greenspace Dynamic WOrld dynamic_world_vegetation_land_percent 2020 40.17

3.3 Percent of vegetation land (Sentinel 2 - NDVI)

Sentinel-2 is a wide-swath, high-resolution, multi-spectral imaging mission supporting Copernicus Land Monitoring studies, including the monitoring of vegetation, soil and water cover, as well as observation of inland waterways and coastal areas. The Sentinel-2 data contain 13 UINT16 spectral bands: four bands at 10 m, six bands at 20 m and three bands at 60 m spatial resolution.

Sentinel-2 images can be used for computing the Normalized Difference Vegetation Index (NDVI) considered as an effective index for estimating green vegetation. The value range of the NDVI is -1 to 1. Negative values of NDVI (values approaching -1) correspond to water. Values close to zero (-0.1 to 0.1) generally correspond to barren areas of rock, sand, or snow. Low, positive values represent shrub and grassland (approximately 0.2 to 0.4), while high values indicate temperate and tropical rainforests (values approaching 1). It is a good proxy for live green vegetation. A threshold of NDVI > 0.4 is used in this analysis for identifying vegetation area based on Sentinel-2 images.

\[NDVI = \frac{NIR(Band 8) - RED(Band 4)}{NIR(Band 8)+ RED(Band 4))}\]

city_id indicator_theme data_sources indicator_name year value
RWA-Kigali greenspace COPERNICUS/S2 s2_ndvi_vegetation_land_percent 2020 87.29
RWA-Kigali greenspace COPERNICUS/S2 s2_ndvi_vegetation_land_percent 2019 88.61
RWA-Kigali greenspace COPERNICUS/S2 s2_ndvi_vegetation_land_percent 2018 90.06
RWA-Kigali greenspace COPERNICUS/S2 s2_ndvi_vegetation_land_percent 2017 86.33
RWA-Kigali greenspace COPERNICUS/S2 s2_ndvi_vegetation_land_percent 2016 88.20

3.4 Percent of land with tree cover (Tree Outside of Forests)

The Tree Outside of Forests data enables the reporting of tree cover within urban areas. Only data corresponding to the year 2020 is available. The proposed indicator measures the average tree cover percent within the selected city.

city_id indicator_theme data_sources indicator_name year value
RWA-Kigali greenspace Tree Outside Forests tof_avg_tree_cover 2020 8.59

3.5 Percent of built area with vegetation (Dynamic World + Sentinel2)

This indicator measures the percentage of vegetation in built areas by combining the Dynamic World land classes to extract built areas and Sentinel-2 imagery to estimate vegetation index based on the NDVI metric.

city_id indicator_theme data_sources indicator_name year value
RWA-Kigali greenspace Dynamic World / Sentinel-2 built_land_cover_with_vegetation_percent 2020 38.87
RWA-Kigali greenspace Dynamic World / Sentinel-2 built_land_cover_with_vegetation_percent 2019 40.81
RWA-Kigali greenspace Dynamic World / Sentinel-2 built_land_cover_with_vegetation_percent 2018 40.73
RWA-Kigali greenspace Dynamic World / Sentinel-2 built_land_cover_with_vegetation_percent 2017 37.06
RWA-Kigali greenspace Dynamic World / Sentinel-2 built_land_cover_with_vegetation_percent 2016 34.17

3.6 Percent of built area with tree cover (Dynamic World + TOF)

This indicator measures the tree cover of built areas by combining the Dynamic World land classesto extract built areas and TOF data for tree cover levels.

city_id indicator_theme data_sources indicator_name year value
RWA-Kigali greenspace Dynamic World / Tree Outside Forests built_land_with_tree_cover_percent 2020 6.947661

3.7 Percent of urban land classes (Intra Urban Land Use)

The ULU data provides land use and land cover information of urban areas based on the application of a supervised classification model trained on high resolution Sentinel-2 satellite imagery data. Urban land classes include:

  • Open space: It refers to any lot of pervious land cover, including farmland.
  • Non residential: This category include commercial and industrial usage,like warehouses, stores, factories, and air- or seaports.
  • Residential: This category include various levels of residential subcategories (atomistic, informal/formal subdivisions and housing projects).

3.8 Percent of tree cover and vegetation by urban land classes (Intra Urban Land Use + Tree Outside of Forests)

This indicator measures the percent of tree cover (based on TOF dataset) and vegetation (based on Sentinel2 imagery) by urban land use classes.

Urban Land Use Class code Urban Land Use Class label Tree cover percent Year City id Vegetation percent
0 open_space 8.98 2020 RWA-Kigali 94.97
1 nonresidential 9.33 2020 RWA-Kigali 27.57
2 atomistic 2.70 2020 RWA-Kigali 17.17
3 informal_subdivision 4.89 2020 RWA-Kigali 29.11
4 formal_subdivision 14.75 2020 RWA-Kigali 37.63
5 housing_project 6.56 2020 RWA-Kigali 32.75